A5063979949- Advanced Vision and Imaging
- Image Enhancement Techniques
- Remote Sensing and LiDAR Applications
- Spectroscopy Techniques in Biomedical and Chemical Research
- Retinal Imaging and Analysis
- Visual Attention and Saliency Detection
- Advanced Energy Technologies and Civil Engineering Innovations
- Structural mechanics and materials
- Material Properties and Failure Mechanisms
- Computer Graphics and Visualization Techniques
- Aerospace and Aviation Technology
- Advanced Image Processing Techniques
Brown University
2024
META Health
2023
Meta (Israel)
2019
Voronezh State University
2016
In order to provide an immersive visual experience, modern displays require head mounting, high image resolution, low latency, as well refresh rate. This poses a challenging computational problem. On the other hand, human system can consume only tiny fraction of this video stream due drastic acuity loss in peripheral vision. Foveated rendering and compression save computations by reducing quality However, cause noticeable artifacts periphery, or, if done conservatively, would modest savings....
One of the urgent tasks development construction associated with new building designs, use which provides increased strength, crack resistance, reducing flow materials, labor intensity, energy consumption and cost. Ensuring effective functioning structures during their operation in harsh environments not only task developing materials higher strength corrosion but also composites as structural resistance is largely determined by ability structure to prevent formation growth cracks. For...
We propose a physically motivated deep learning framework to solve general version of the challenging indoor lighting estimation problem. Given single LDR image with depth map, our method predicts spatially consistent at any given position. Particularly, when input is an video sequence, not only progressively refines prediction as it sees more regions, but also preserves temporal consistency by keeping refinement smooth. Our reconstructs spherical Gaussian volume (SGLV) through tailored 3D...
Recent advances in head-mounted displays (HMDs) provide new levels of immersion by delivering imagery straight to human eyes. The high spatial and temporal resolution requirements these pose a tremendous challenge for real-time rendering video compression. Since the eyes rapidly decrease acuity with increasing eccentricity, providing peripheral vision is unnecessary. Upcoming VR estimation gaze, enabling gaze-contingent compression methods that take advantage this falloff. In setting,...
Gaussian splatting methods are emerging as a popular approach for converting multi-view image data into scene representations that allow view synthesis. In particular, there is interest in enabling synthesis dynamic scenes using only monocular input -- an ill-posed and challenging problem. The fast pace of work this area has produced multiple simultaneous papers claim to best, which cannot all be true. work, we organize, benchmark, analyze many Gaussian-splatting-based methods, providing...
We propose a physically-motivated deep learning framework to solve general version of the challenging indoor lighting estimation problem. Given single LDR image with depth map, our method predicts spatially consistent at any given position. Particularly, when input is an video sequence, not only progressively refines prediction as it sees more regions, but also preserves temporal consistency by keeping refinement smooth. Our reconstructs spherical Gaussian volume (SGLV) through tailored 3D...